The effect of surgical approach on performance of lymphadenectomy and perioperative morbidity for radical nephroureterectomy.

OBJECTIVES

To examine the effect of surgical approach on regional lymphadenectomy (LND) performance and inpatient complications for radical nephroureterectomy (NU) using a national administrative database.

METHODS

The National Inpatient Sample (2009-2012) was used to identify patients who underwent NU for urothelial carcinoma. Cohorts were stratified by performance of LND. Covariates included patient demographics, comorbidity, hospital characteristics, hospital volume, performance of LND, surgical approach (open [ONU], laparoscopic [LNU], or robotic [RNU]), and complications. Multivariable logistic regression was used to identify factors associated with LND performance and complications.

RESULTS

A weighted population of 14,059 (85%) without LND and 2,560 (15%) with LND was identified. LND was more common in RNU (27%) compared with ONU (15%) and LNU (10%) (P<0.01). On multivariable analysis, when compared with ONU, RNU was associated with increased odds of LND performance (odds ratio [OR] = 1.9, 95% CI: [1.3-2.8]; P = 0.001), whereas LNU was associated with decreased odds of LND performance (OR = 0.6, 95% CI: [0.4-0.8]; P = 0.004). Multivariable analysis of risk factors for complications demonstrated lower odds of complications with RNU (OR = 0.6, 95% CI: [0.4-0.8]; P = 0.001), whereas performance of LND increased the risk of complications (OR = 1.3, 95% CI: [1.001-1.7]; P = 0.049).

CONCLUSIONS

When compared with ONU, RNU increased the odds of LND performance and had a lower inpatient complication rate, whereas LNU reduced the odds of LND performance and had no significant effect on inpatient complication rates. Performance of LND was independently associated with higher inpatient complication rates.